Introduction
While carrying out a research, it is very important to determine the reliability and viability of the research results. While reliability emphasizes on the consistency of the research results, validity emphasizes on investigating the exact item that the researcher aims at examining. Essentially, validity helps researchers to determine the feasibility of any presumed inferences. In this case, the researcher aims at determining whether fun in the workplace can influence employees’ motivation, job satisfaction, collaboration, and general performance of the organization. In the process of carrying out the research, it is imperative to evaluate the external, internal, and construct validity of the research methodology.
Contrasts of external, internal, and construct validity
To attain construct validity, researchers rely on an existing presumption to develop a research topic (Bambale, Shamsudin, & Subramaniam, 2013). In this case, the presumption is that fun in the workplace motivates employees, and it increases their satisfaction levels. The conjecture prompts the researcher to find out whether various activities that bring fun in the workplace can indeed influence the organizations’ performance. Construct validity defines the presumptions that the researcher seeks to investigate. It requires the researcher to use precise definitions to describe an assumption without necessarily interpreting the research results.
Appropriate representation of the results enables the researcher to achieve construct validity with ease. From one time to another, construct validity obligates the researcher to define the study topic and make meaningful projected outcomes. The researcher has to provide evidence regarding the variables of study based on prior theory. In this case, the researcher will achieve construct validity by providing theoretical evidence that fun in the work place enhances organizational performance.
External validity embarks on generalization, where, the researcher validates the results obtained from the sampled population by determining their correctness in another study (Ferguson, 2004). In studying how fun in the workplace influences employees’ productivity, the study results would be externally valid if they were valid in other places. This means that if the researcher was to carry out a similar research elsewhere at a later date, the study results would be somewhat related.
Therefore, if the research results enable the researcher to generalize the inferences, the study becomes externally valid. Sampling is an imperative exercise in obtaining a good sample to represent a certain population. The researcher should develop an equitable sample, which can generalize the population of study. In case it is impossible to develop the equitable sample, the researcher can employ the proximal similarity model. The model enables the researcher to obtain a sample using similarities of the region, people, and period of study.
Internal validity seeks to identify the dependent and the independent variables, and determine the extent to which the independent variables influence the dependent variables. Internal validity seeks to measure the extent to which the researcher can attribute the research results to the experimental treatments (Creswell, 2009). In this case, the researcher must have the confidence to state that various activities that bring fun in the workplace create some job satisfaction, which results into increased employee productivity. The researcher may have to measure the productivity of employees subjected under different treatments to ascertain that the outcomes result from the experimental treatments.
Causality is one of the factors of internal validity that measure the effects that particular variables may have on the measurable outcomes. Causality discards the presumption that similar changes to a variable may have similar outcomes. It underpins that a certain outcome signifies that a change might play a great role in influencing the dependent variable. In this case, an increase in the organizations’ performance would be a clear indication that adopting a new way of bringing fun in the workplace can enhance the productivity of the employees. Temporal precedence obligates the researcher to demonstrate that the cause happened before the effect. The cause and effect must relate in some way, and in case there is no feasible effect, there ought to be a plausible alternative explanation.
Comparing external, internal, and construct validity
From the descriptions of external, internal, and construct validity, it is evident that they are all concerned with the confidence that a researcher will have regarding certain research results. It is worth noting that external validity and construct validity emphasize on generalization. In both cases, researchers will find validity if the propositions and the inferences that they make are valid elsewhere. The three methods of validating research results outline the necessity of studying a sample that represents the population of interest. The sample produces the best results if it is a good representation of the population.
Dependent and independent variables are present in the three methods of measuring validity, where, a change in the independent variable is likely to cause a change in the dependent variable (Cozby, 2012). Although internal validity does not embark on generalization, the cause and effect characteristic relates it to external and construct validity. Construct validity will always embark on accurate methods to analyze variables, and similar to internal validity, it highlights on the extent to which the researcher can defend the study results. Researchers can always attain internal, external, and construct validity by designing the research appropriately.
Threats to external validity
- Testing: Pre-tests expose the respondents to the experimental treatments; therefore, it would not be feasible to generalize the results of a pretested population. Essentially, pre-tests give respondents a clue about the study topic, and their responses may differ from those of the respondents without a clue about the study topic.
- Selection bias: When the study sample is not a good representation of the population of study, the research results are unfit for generalization.
- Hawthorne effect: When respondents are aware that they will take part in a study, their anxiety is likely to influence their responses, and the results would be unfit for generalization.
- Multi-treatment interference may occur if the same respondents receive similar treatments repeatedly. In such a case, it would be unfeasible to generalize the results for a different population, which may receive single treatments.
Threats to construct validity
- Construct validity relies on a universal cohesion; therefore, an erroneous construct can invalidate an entire experiment (Franklin, 2012).
- Mono-operation bias may occur if the researcher relies on one variable to make deductions about an experiment.
- Interaction effect may occur if the researcher resolves to influence some participants to respond in a certain manner.
- Hypothesis guessing and researcher’s expectations may cause prejudice. In this case, the researcher may look forward to some results and formulate subjective goals that may create unreliable results.
Impact of validity issues on the envisioned research
The envisioned research will establish the correlation between fun in the workplace and organizational performance. The researcher will establish internal validity by establishing the causality between employees’ motivation, job satisfaction, collaboration, and their productivity levels. There is a possibility of having a direct relationship between employees’ productivity and fun in the work place. The envisioned research will consider external validity by taking strict measures while selecting the sample to take part in the study. The sampled population must comprise of different types of employees from various workplaces.
The researcher will ensure equitability by ensuring that the sample comprises of employees of different genders and different ranks in organizations. It will be imperative to give a clear definition of the similarities and differences of sampled participants in terms of population, place, and time, and determine how well they represent the entire population. Adequate definitions, measures of variables, and preexisting theories about fun in the workplace and organization’s productivity will play a great role in achieving construct validity. The study will have to show that fun in the workplace is highly correlated with organizational productivity to maintain construct validity.
Conclusion
From the discussions, it is evident that internal, external, and construct validity are imperative in any research. A careful choice of the sampling method would be necessary to minimize the threats to validity. In this case, a random sampling method would work perfectly in realizing generalization. The researcher can lessen the threats to external validity by giving a clear description of the items of study. It would be vital to compare and contrast the people, places, and the period of study to achieve generalization. Overall, researchers should not attempt to predict outcomes while carrying out any research; they should be ready to accept any outcomes to provide reliable and viable results.
References
Bambale, A., Shamsudin, F., & Subramaniam, C. (2013). The construct validity of servant leadership in public utility organizations. International Journal of Global Business, 6(2), 16-33.
Cozby, P. C. (2012). Methods in behavioral research. Boston: McGraw-Hill Higher Education.
Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods approaches. Thousand Oaks, CA: Sage Publications.
Ferguson, L. (2004). External validity, generalizability, and knowledge utilization. Journal of Nursing Scholarship, 36(1), 16-22.
Franklin, M. I. (2012). Understanding research: Coping with the quantitative-qualitative divide. London: Routledge.